Underwater Image Page 1

Underwater Image Quality Measurement and

Improvement

Report submitted for the partial fulfillment of the requirements for the degree of

Bachelor of Technology in

Information Technology

Submitted by

SUMANA MAJUMDER (IT2014/034)

NISHA DUTTA (IT2014/041)

PRIYANKA MONDAL (IT2014/043)

Under the Guidance of Mr. Hiranmoy Roy

[Asst. Prof. Department of IT]

RCC Institute of Information Technology

Canal South Road, Beliaghata, Kolkata – 700 015

[Affiliated to West Bengal University of Technology]

Underwater Image Page 2

Acknowledgement

We would like to express our sincere gratitude to Mr. Hiranmoy Roy of the department of

Information Technology, whose role as project guide was invaluable for the project. We are

extremely thankful for the keen interest he took in advising us, for the books and reference

materials provided for the moral support extended to us.

Last but not the least we convey our gratitude to all the teachers for providing us the technical

skill that will always remain as our asset and to all non-teaching staff for the gracious hospitality

they offered us.

Place: RCCIIT, Kolkata

Date:

SUMANA MAJUMDER

NISHA DUTTA

PRIYANKA MONDAL

Underwater Image Page 3

Department of Information Technology

RCCIIT, Beliaghata,

Kolkata – 700 015,

West Bengal, India

Approval

This is to certify that the project report entitled “Underwater image quality measurement and

Improvement” prepared under my supervision by Sumana Majumder (IT2014/034), Nisha Dutta

(IT2014/041), Priyanka Mondal (IT2014/043),be accepted in partial fulfillment for the degree of

Bachelor of Technology in Information Technology.

It is to be understood that by this approval, the undersigned does not necessarily endorse or

approve any statement made, opinion expressed or conclusion drawn thereof, but approves the

report only for the purpose for which it has been submitted.

……………………………………….. … ………………………………………

Dr. Abhijit Das Hiranmoy Roy

Assoc. Prof. & HOD Asst. Prof.

Department of IT Department of IT

RCCIIT RCCIIT

Underwater Image Page 4

Contents

1. Introduction

2. Problem Definition

3. Literature Survey

4. SRS (Software Requirement

Specification)

5. Planning

6. Design

7. Results

8. List of Tables

9. List of Figures

10.References / Bibliography

Page Numbers

5

6

7

13

14

15

16

16

18

20

Underwater Image Page 5

INTRODUCTION

The quality of underwater images plays a pivotal role in ocean engineering and scientific

research, such as monitoring sea life, accessing geological environment, and ocean rescue.

However, the absorption and scattering effects of the water limit the visibility of the underwater

objects. Consequently, the images captured by underwater cameras usually suffer from low

contrast, no uniform illumination, blurring, bright artifacts, diminished color, noise, and other

distortions. Many algorithms have been proposed for restoring colors and enhancing contrast for

observed underwater images. In the majority of these methods, visual inspection is used for

evaluating the image processing algorithms’ performances. However, subjective evaluation is

biased, and it is expensive with respect to time and resources. More importantly, subjective

evaluation cannot be automated. Therefore, it is important to have a reliable objective evaluation

measure. Generally, objective evaluation methods can be classified into two categories: no

reference measures and full-reference measures, depending on whether the images with the “true

color” and “ideal contrast” are available. In underwater image processing scenarios, such ideal

images are usually unavailable. Therefore, the non reference image quality measures are desired.

This challenge is further exacerbated by the need to develop an underwater image quality

measure that captures the objectivity and perception of the human visual system (HVS).

Contrast enhancement technique is widely used for underwater image processing to

improve the contrast performance. The development of the contrast enhancement technique for

underwater image has attracted considerable attention in recent years. Researchers are improving

the image contrast to extract as many information as possible by applying various algorithms.

Underwater Image Page 6

PROBLEM DEFINATION

Underwater images suffer from blurring effects, low contrast, and grayed out colors due

to the absorption and scattering effects under the water. Few image segmentation algorithms

have been developed for that image segmentation. Unfortunately, no well-accepted mechanism

exists that can segment the underwater images similar to human perception. A lot of fuzziness is

present in those images. Therefore, fuzzy concept based segmentation is more useful in that

situation. A human visual system concepts based segmentation will be more accurate. To address

the problem, a new non reference underwater image quality measure (UIQM) is presented. The

UIQM comprises three underwater image attribute measures: the underwater image

colourfulness measure (UICM), the underwater image sharpness measure (UISM), and the

underwater image contrast measure (UIConM).

The quality of underwater image is poor due to the properties of water and its impurities.

The properties of water cause attenuation of light travels through the water medium, resulting in

low contrast, blur, inhomogeneous lighting, and color diminishing of the underwater images.

Here a method of enhancing the quality of underwater image is proposed. The proposed method

consists of two stages: 1) Modified Von Kries hypothesis and 2) Global histogram stretching.

Underwater Image Page 7

LITERATURE SURVEY

From “Human-Visual-System-Inspired Underwater Image Quality Measurement” by Karen

Panetta et al., we have learned about image quality measurement. Measurements of an

underwater image are UICM (Underwater Image Colorfulness Measure), Underwater Image

Sharpness Measure (UISM), Underwater Image Contrast Measure (UIConM).

A. Underwater Image Colorfulness Measure (UICM)

Many underwater images suffer from a severe color-casting problem. As the depth of the water

increases, colors attenuate one by one depending on their wavelength. The color red disappears

first due to it possessing the shortest wavelength. As a result, underwater images usually

demonstrate a bluish or greenish appearance. Furthermore, limited lighting conditions also

causes severe color de-saturation in underwater images. A good underwater image enhancement

algorithm should produce good color rendition. The HVS captures colors in the opponent color

plane. Therefore, the two opponent color components related with chrominance RG and YB are

used in the UICM, as shown in

(1)

(2)

Underwater images usually suffer from heavy noise. Therefore, instead of using the regular

statistical values, the asymmetric alpha-trimmed statistical values are used for measuring

underwater image colorfulness. The mean is defined by:

R

L

TK

Ti

iRG

RL

RG IntensityTTK

1

,,

1 (3)

The second-order statistic variance in:

2

,

1

,,2 )(

1RG

N

p

pRGRG IntensityN

(4)

BGR

YB

GRRG

2

Underwater Image Page 8

The overall colourfulness metric used for measuring underwater image colourfulness is

demonstrated in

YBRGYBRGUICM ,2

,2

,2

,2 1586.00268.0 (5)

B. Underwater Image Sharpness Measure (UISM)

Sharpness is the attribute related to the preservation of fine details and edges. For images

captured under the water, severe blurring occurs due to the forward scattering. This blurring

effect causes degradation of image sharpness. To measure the sharpness on edges, the Sobel edge

detector is first applied on each RGB color component. The resultant edge map is then multiplied

with the original image to get the grayscale edge map. By doing this, only the pixels on the edges

from the original underwater image are preserved. It is known that the enhancement measure

estimation (EME) measure is suitable for images with uniform background and shown non

periodic patterns accordingly; the EME measure is used to measure the sharpness of edges.

The UISM is formulated as shown in

3

1

)(c

cc edgegrayscaleEMEUISM (1)

EME=

1

1

2

1 ,min,

,max,)log(

21

2 k

l

k

k lk

lk

I

I

kk (2)

C. Underwater Image Contrast Measure (UIConM)

Contrast has been shown to correspond to underwater visual performance such as stereoscopic

acuity. For underwater images, contrast degradation is usually caused by backward scattering.

The contrast is measured by applying the logAMEE measure on the intensity image as shown in

)(log IntensityAMEEUIConM

The logAMEE in

)log(21

1log

,min,,max,

,min,,max,1

1

2

1 ,min,,max,

,min,,max,

lklk

lklkk

l

k

k lklk

lklk

II

II

II

II

kkAMEE

(1)

Underwater Image Page 9

Underwater image quality enhancement through Rayleigh-

stretching

A. Modified of Von Kries hypothesis

Image channels are applied with modified Von Kries hypothesis. First of all, the average values

of each channel, Ravg, Gavg, Bavg are calculated

M

i

N

j

Bavg

M

i

N

j

Gavg

M

i

N

j

Ravg

jiINM

B

jiINM

G

jiINM

R

1 1

1 1

1 1

),(1

),(1

),(1

(2)

M x N indicates the number of pixels in a channel and is the pixel value of respective

channel at position of (i,j).

The median value is determined from these three average values and used as the

reference or target value. The remaining color channels are determined with multipliers

(A and B) in order to produce a balanced image. The equations are used to calculate the

multipliers for maximum and minimum average values, respectively.

(4)

(5)

Based on the calculation in equations, the color channel with a minimum intensity value is

multiplied with multiplier A whereas the color channel with a maximum intensity value is

multiplied with multiplier B.

(1)

(3)

),,max(

),,(

),,min(

),,(

avgavgavg

avgavgavg

avgavgavg

avgavgavg

BGR

BGRmedianB

BGR

BGRmedianA

Underwater Image Page 10

B. Global histogram stretching

In order to spread the pixels values of the image, the histogram of the image channels are applied

with global stretching, where the histogram are stretched over the whole dynamic range of [0,

255]. This is also as preparation of the next step where the histogram will be divided into two

regions based on its average value. The stretched-histogram will provide a better pixel

distribution of the image channels and thus gives a more accurate average value of the channel

which represents the average value of the channel for the whole dynamic range.

The following equation is used to stretch the histogram of respective color channel to the whole

dynamic range.

min

minmax

minmaxmin ))(( o

ii

ooiPP inout

(6)

Pin and Pout are the input and output pixels, respectively, and imin, imax, omin, and omax are

the minimum and maximum intensity level values for the input and output images, respectively.

C. Division and stretching of image histogram with respect to Rayleigh distribution

The histogram is then divided into two regions based on this average value. By dividing the

histogram at its average point, two regions are produced namely lower and upper regions. The

lower region should have the intensity range between 0 to the average value of the image

histogram whereas the upper region should have the intensity range between the average value to

the maximum intensity range of 255.

For the next step, these two regions of histogram are stretched independently to produce

two separated histograms. In addition to this step, these regions are stretched to follow the

Rayleigh distribution over the entire dynamic range of [0, 255]. The lower region which has the

intensity value between 0 to the average value will be stretched to the entire dynamic range of [0,

255]. The same process is applied to the upper region, where the initial range of the region,

which lies from average value to 255 is stretched to the entire dynamic range of [0, 255].

The probability distribution function of Rayleigh distribution is given by the equation

2

2

2

2

x

Rayleigh ex

PDF for 0x , 0 (7)

where α is the distribution parameter of Rayleigh distribution and x is the input data which is, in

this case, the intensity value.

Rayleigh-stretched distribution in equation

Underwater Image Page 11

2

2

minminmax

minmaxmin

2

2

min

minmax

minmaxmin

..

oii

ooiP

inin

e

oii

ooiP

stretchedRayl (8)

The output histogram is stretched over the entire dynamic range of [0, 255]. Therefore,

the values of omax and omin can be substituted with the values of 255 and 0, respectively. Thus,

previous equation can be simplified as below equation:

2minmax2

2min

2

255

minmax

2

min .255

.ii

iP

in

in

eii

iPstretchedRayl

Underwater Image Page 12

where imin and imax indicate the minimum and maximum intensity level values for input image in

each region, respectively. For the lower region, imin indicates the minimum intensity level value

of the image histogram, and imax indicates the maximum intensity level value in the region, which

is equivalent to the average-value of the histogram. For the upper region, imin indicates the

minimum intensity level value in the region which is equivalent to the average-value of the

image histogram, and imax indicates the maximum intensity level value in the region which is

equivalent to the maximum intensity level of the image.

Underwater Image Page 13

SRS (Software Requirement Specification)

Operating System Processor Disk Space RAM

Windows 10 Any Intel or AMD x

86-64 processor

2GB for MATLAB

only

2GB

Underwater Image Page 14

PLANNING

Software Life Cycle Model which one we have used is Iterative Waterfall Model

Underwater Image Page 15

DESIGN

Underwater Image Page 16

Results

Quality Measurement Table:

Images Underwater Image Quality Measure

UICM UISM UIConM

Image 1 -3.3947 0.5940 -6.4732

Image 2 -5.6569 3.0707 -8.8825

Image 3 -4.6389 4.7524 -9.4221

Image 4 -8.2311 5.2647 -9.587

Image 5 -4.7714 5.2405 -9.3524

Image 6 -2.7894 3.0245 -7.2689

Image 7 -3.3154 1.0780 -10.3718

Underwater Image Page 17

GUI Design:

Underwater Image Page 18

LIST OF FIGURES

(a) (b) (c)

(d) (e) (f)

(g)

Underwater Image Page 19

Image Improvement:

Original Image Output Image

Original Image Output Image

Original Image Output Image

Underwater Image Page 20

REFERENCES / BIBLIOGRAPHY

1. K. Panetta, S. Agaian, Y. Zhou, and E. J. Wharton, “Parameterized logarithmic

framework for image enhancement,” IEEE Trans. Syst. Man Cybern. B, Cybern., vol. 41,

no. 2, pp. 460–473, Apr. 2011.

2. Karen Panetta, Fellow, IEEE, Chen Gao, Student Member, IEEE, and Sos Agaian, Senior

Member, IEEE, “IEEE JOURNAL OF OCEANIC ENGINEERING”, VOL. 41, NO. 3,

541 Human-Visual-System-Inspired Underwater Image Quality Measures , JULY 2016.

3. Miao Yang, Member, IEEE, and Arcot Sowmya, Member, IEEE, “An Underwater Color

Image Quality Evaluation Metric”, IEEE TRANSACTIONS ON IMAGE

PROCESSING, VOL. 24, NO. 12, DECEMBER 2015.

4. Ahmad Shahrizan Abdul Ghani and Nor Ashidi Mat Isa,” Underwater image quality

enhancement through composition of dual-intensity images and Rayleigh-stretching”,

Abdul Ghani and Mat Isa SpringerPlus 2014, 3:757.

5. Arici T, Dikbas S, Altunbasak Y (2009) A histogram modification framework and its

application for image contrast enhancement. IEEE Trans Image Process 18:No. 9.

6. Barnard K, Cardei V, Funt B (2002) A comparison of computational color constancy

algorithms-part I: experiments with image data. IEEE Trans Image Process Vol. II:No. 9.

7. Bassiou N, Kotropoulos C (2007) Color image histogram equalization by absolute

discounting back-off. Comput Vis Image Und 107(1–2):108–122.

8. Church SC, White EM, Partridge UC (2003) Ultraviolet dermal reflection and mate

choice in the guppy. Elsevier Science Ltd, pp 693–700.

9. Eustice R, Pizarro O, Singh H, Howland J (2002) Underwater image toolbox for optical

image processing and mosaicking in MATLAB. In: Proceedings international symposium

on underwater technology., pp 141-145.

Underwater Image Page 21

10. Gasparini F, Schettini R (2003) Color correction for digital photographs. In: 12th

International Conference on Image Analysis and Processing (ICIAP 2003)., pp 646-651.

11. Hitam MS, Yussof WNJW, Awalludin EA, Bachok Z (2013) Mixture contrast limited

adaptive histogram equalization for underwater image enhancement. In: IEEE

International Conference on Computer Applications Technology (ICCAT), Sousse, pp 1–

5.

12. Iqbal K, Salam RA, Osman A, Talib AZ (2007) Underwater image enhancement using

integrated color model. IAENG Int J Comput Sci 34:2 IJCS 34-2-12.

13. Iqbal K, Odetayo M, James A, Salam RA, Talib AZ (2010) Enhancing the low quality

images using unsupervised color correction method. In: International conference on

system man and cybernatics (SMC), Istanbul, pp 1703–1709.

14. Jaffe JS (1990) Computer modeling and the design of optimal underwater imaging

systems. IEEE J Ocean Eng 15(2):101–111.

15. Kim Y (1997) Contrast enhancement using brightness preserving bi-histogram

equalization. IEEE Trans Consum Electr 43(1):1–8.

16. Kumar R, Rattan M (2012) Analysis of various quality metrics for medical image

processing. IJARCSSE 2:11–160.

17. McGlamery BL (1979) A computer model for underwater camera systems. Proc SPIE

208:221–231.

18. Naim MJNM, Mat Isa NA (2012) Pixel distribution shifting color correction for digital

color images. Applied Soft Computing 12(Issue 9):2948–2962.

19. Padmavathi G, Muthukumar M, Thakur SK (2010) Non Linear Image Segmentation

using Fuzzy C-means Clustering Method with Thresholding for underwater images.

IJCSI 7(3):35.

Underwater Image Page 22

20. Rizzi A, Gatta C, Marini D (2003) A new algorithm for unsupervised global and local

color correction. Pattern Recogn Lett 24:No. 11:1663–1677.